Data Sheet 1_Optimizing healthcare resources in pyogenic liver abscess: a dual-threshold H

138 adult patients from a prospective tertiary-center cohort (2018–2023) form the core of this dataset, which also includes 38 ICU patients from MIMIC-IV and 9,226 individuals from NHANES 2017–2020 for validation. Mingzhu Tao created this dataset to examine how admission biomarkers like HDL-C and CRP predict length of stay for pyogenic liver abscess. The dataset was last updated on April 30, 2026.

Use Cases

• Predicting length of hospital stay based on admission HDL-C and CRP levels.
• Validating a dual-threshold risk model (HDL-C < 1.03 mmol/L and CRP > 1.0 mg/dL) for identifying high-risk patients.
• Analyzing the mediating role of C-reactive protein in the relationship between HDL-C and hospitalization duration.
• Building multivariable models that incorporate abscess size and sepsis status for patient outcome prediction.

Strengths

• Includes data from three distinct cohorts totaling over 9,400 individuals for robust validation.
• Prospective cohort data from a tertiary center spans 5 years (2018–2023).
• Model performance metrics are reported, including an R² of approximately 0.66 and RMSE of about 6.4 days.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the primary cohort is known (n=138), but the total structure of the combined dataset is unclear.
• Data may reflect bias inherent to the specific hospital and ICU settings used.

Provenance

Source

figshare, authored by Mingzhu Tao.

Collection Method

Analysis of a prospective adult PLA cohort, validated with MIMIC-IV ICU and NHANES data.

Time Range

Primary cohort: 2018–2023; NHANES validation: 2017–2020.

Freshness

Last updated 2026-04-30 04:16:24; freshness should be verified.